5-Lumps kinetic modeling, simulation and optimization for hydrotreating of atmospheric crude oil residue

Abstract

This research aims at developing a discrete kinetic model of the hydrotreating process of a crude oil residue based on experiments. Thus, various experiments were conducted in a continuous flow trickle bed reactor over a temperature range of 653-693 K, liquid hourly space velocity of 0.3-1.0 h -1 , and hydrogen pressure of 6.0-10.0 MPas at a constant hydrogen to oil ratio of 1000 L L -1 . The reduced crude residue had been assumed to have five lumps: naphtha, kerosene, light gas oil, heavy gas oil and vacuum residue. An optimization technique based on the minimization of the sum of the squared error between the experimental and predicted compositions of the distillate fractions was used to calculate the optimal value of kinetic parameters. The predicted product composition showed good agreement with the experimental data for a wide range of operating conditions with a sum of square errors of less than 5 %.

Accelerating Kinetic Parameter Identification by Extracting Information from Transient Data: A Hydroprocessing Study Case

“…To be able to compare the prediction accuracy of both modeling strategies, the quality performance was calculated based on 38 steady-state points of 38 episodes. The used metrics of the MAPE and RMSE formula are given in Equations (8) and (9). As liquid product nitrogen in the experimental data varies across a wide range of values (0.6-500 ppm), it is preferable to simultaneously analyze MAPE and RMSE to compare the prediction accuracy.…”

“…Hydroprocessing kinetic modeling is a challenging and time-consuming task as the crude oils contain many compounds with complicated structures [6]. Diverse approaches for hydroprocessing modeling have been developed in the past such as the lumping technique [7][8][9][10][11][12], detailed kinetic modeling [13][14][15][16][17][18][19], and black-box approach [20][21][22][23]. The lumping technique, which consists of regrouping chemical compounds with similar properties, is the most common and widely used [24].…”

Asymmetric Probability Distribution Function-Based Distillation Curve Reconstruction and Feature Extraction for Industrial Oil-Refining Processes

“…For more details about the distillation process, refer to the previous literature. − Obviously, it is the reaction system rather than the fractionation system that determines the distribution of end products primarily. The dynamic modeling for the crucial reaction system has received more and more attention, such as the classical Stangeland model, the product-oriented model, the discrete lumped models, , and the continuous lumped models. ,,− Among the existing methods, continuous lumping has become extremely popular in recent years. With the distillation curves of inlet reactants and outlet products, it is not difficult to set up equilibrium equations for the reaction system using dynamics and thermodynamics laws.…”

Stabilization time modeling for hydroprocessing: Identification of the dominant factors

“…Since crude oil contains a lot of compounds with difficult structures, kinetic modeling of hydroprocessing is a challenging and time-consuming task (Jarullah et al, 2011). There are various approaches for hydroprocessing modeling, such as the detailed kinetic (Schweitzer et al, 1999;Oyekunle and Edafe, 2009;Charon-Revellin et al, 2011;Nguyen et al, 2015;Raghuveer et al, 2016;Nguyen et al, 2017;Doukeh et al, 2018), lumping modeling (Bonnardot, 1998;López García et al, 2010;Lababidi and AlHumaidan, 2011;Tang et al, 2013;Becker et al, 2015;Esmaeel et al, 2016) and black-box approach (Elkamel et al, 1999;Bahmani et al, 2010;Sadighi and Reza Zahedi, 2013;Sadighi and Mohaddecy, 2018). The most common and widely used is the lumping approach which consists of regrouping chemical compounds with similar properties (Oliveira et al, 2016).…”